The present invention is directed to an error diffusion process which allows for elimination of isolated subpixels before rendering by a printer or printing system. More specifically, the present invention is directed to an error diffusion process for eliminating the creation of isolated subpixels and adjusting the diffused error value so as to compensate for the elimination of the isolated subpixels and localized error.
Error diffusion is a common technique for converting a grey scale image to a binary image. This process, however, assumes that a printer is an ideal device wherein black pixels and white pixels can be rendered not withstanding their effective size. FIG. 1 shows the block diagram of a conventional error diffusion process.
As illustrated in FIG. 22, input grey video is inputted to an adder 10 wherein slowscan error, which represents error from the processing of the previous scanline of pixels, stored in a FIFO 11 is added to the input grey video. Moreover, fastscan error from an error distribution circuit 15 is also added to the input grey video at adder 10. The fastscan error from the error distribution circuit 15 represents the error from processing the previous pixel in the same scanline. The modified input grey video (PixN) is then fed to a comparator 14 which compares the modified input grey video with a threshold value. Based on the comparison with the threshold value, the comparator 14 outputs a binary output of either 1 or 0. The modified input grey video is also fed to a subtraction circuit 12 and a multiplexer 14. Subtraction circuit 12 generates a value representative of the difference between a black reference value and the modified input grey video value. This difference is also fed to multiplexer 14. Multiplexer 14 selects either the difference value or the modified input grey video value as the pixel error for the presently processed pixel based on the binary output from comparator 14. This pixel error is fed to the error distribution circuit 15 which utilizes a plurality of weighting coefficients to distribute the error to various adjacent pixels.
However, with the recent improvements in the capabilities of printers, conventional error diffusion cannot be readily used without experiencing artifacts in the rendered image. For example, many printers now use high addressable outputs; two or more binary bits are generated for each grey pixel input. Usually, the multiple bits are created in the fastscan direction (the orientation in which the single scanline is printed).
High addressability is important in situations where the device can process the image data at one resolution, but print at a higher resolution. In such a situation, the present invention can take advantage of a processing system designed for a lower resolution image, (lower resolution can be processed quicker and less expensively), and a printing device which, through laser pulse manipulation, can print at a higher resolution. For example, the image can be processed at 600xc3x97600xc3x978 and printed at 2400xc3x97600xc3x971 using the high addressability process of the present invention. In the above example, the high addressability characteristic is 4. If the image was processed at 600xc3x97600xc3x978 and printed at 1200xc3x97600xc3x971, the high addressability characteristic would be 2.
In such a high addressable environment, conventional error diffusion process can generate images that contain many isolated subpixels. An isolated subpixel is a subpixel that is different from both of it""s neighbors in the fastscan direction; i.e., a black subpixel surrounded by white subpixels. At first blush this would not seem to be a problem, but xerography is not sensitive enough to effectively print single isolated subpixels, thus resulting in objectionable artifacts being created in the rendered image.
One such artifact that is caused by the inability of a xerographic system to render a subpixel is a grey level shift in the output data. More specifically, the grey level shift is caused because the isolated subpixels that don""t print, due to the insensitivity of a xerographic printer, do not add to the light absorption as expected and thus the actual grey level perceived is not equal to the grey level of the original image.
For example, if a grey sweep is printed using a high addressability characteristic that is greater than 1, for example 2, the image should appear as a smooth gradient of grey from grey to light grey to white. However, if such a grey sweep is printed utilizing conventional error diffusion and a high addressability characteristic greater than 1, a discontinuity appears in the image near the darker end. This discontinuity is due to the fact that a certain grey level may produce relatively few isolated subpixels, but the adjacent grey levels may produce many more isolated subpixels. The areas with a large percentage of isolated subpixels appear much lighter since the subpixels do not faithfully reproduce.
Another artifact of the inability to render isolated subpixels is that certain grey levels may have whited out areas. This artifact is caused by many isolated subpixels being printed in a localized area. In other words, since the isolated pixels cannot be effectively rendered by the printer, these isolated pixels become white areas in the generated output document. Thus, a grey area may become completely white if the many isolated subpixels are not properly rendered by the printer.
Thus, the present invention proposes a system which compensates for a printer""s inability to render isolated subpixels when using high addressability error diffusion to process the image data, by eliminating the isolated subpixels. The present invention also proposes updating the error propagated in the error diffusion process to account for modifications in the subpixel datastream.
Moreover, the present invention improves the image quality of images processed using subpixel elimination in an error diffusion process. More specifically, the present invention reduces the magnitude of the error signal which is calculated and propagated to pixels in the slow-scan direction after subpixel elimination. In other words, the present invention reduces the variance of the modified error signal but keep overall sum of error unchanged.
One aspect of the present invention is a method for processing image data. The method receives a plurality of pixels of image data; converts the received pixels of image data to a bitmap of image data ready for rendering on an output device; analyzes the bitmap for a predetermined characteristic; modifies the image data within the bitmap when the predetermined characteristic is detected; generates an error based on the modification performed; determines if other image data within the bitmap needs modification based on the generated error; and further modifies the bitmap if other image data within the bitmap needs modification based on the generated error.
A second aspect of the present invention is a system for processing image data. The system includes means for converting received pixels of image data to a bitmap of image data ready for rendering on an output device; comparing means for analyzing the bitmap for a predetermined characteristic; compensation means for modifying the image data within the bitmap when the predetermined characteristic is detected; error means for generating an error based on the modification performed by said compensation means; and local error means for determining if other image data within the bitmap needs modification based on the generated error and for further modifying the bitmap if the other image data within the bitmap needs modification based on the generated error.
Further objects and advantages of the present invention will become apparent from the following descriptions of the various embodiments and characteristic features of the present invention.